An exercise in dexterity

INSTITUTIONAL outcomes from this year's Research Assessment Exercise show the importance of being able to work the system.

Universities that used such skills, in particular by excluding non-research producers, were among the better placed in the league table on page xvi of our pull-out results supplement.

This should not be taken to mean that their research has not improved as well, but there is little doubt that some institutions have suffered in the past by including a long tail of non-producers. This applied equally at either end of the table, at institutions such as Oxford and Napier.

Oxford has been hampered in previous RAEs by taking the view that almost every academic on a contract should be included whether or not they were research productive. Last time it put in 99.9 per cent of its academic staff. This time the figure was 91 per cent.

It was rewarded by a weighted average that places it at the top of our institutional table, displacing Cambridge - top of the two previous RAEs conducted on numerical ratings. Cambridge, however, will dispute any claim that it has lost its standing as Britain's leading research university, having submitted 98 per cent of its academic staff, the same level as in 1992.

Napier, among the new universities subjected to the 1992 exercise within a year of attaining its status, has both learnt from its experience and generated some quality research. In 1992, it put in 29 per cent of its staff and scarcely troubled the scorers. Only the similarly recent University of Glamorgan scored a lower weighted grade average. This time Napier put in 17 per cent of its staff. The bulk attained 3b or 3a ratings, with a small group scoring a 4. It rose to 69th out of 102 in the institutional league table, and to ninth among the new universities.

Other institutions whose rising research standards were buttressed by astutely targeted submissions include Nottingham Trent, Sunderland, Liverpool John Moores, Bristol, Cardiff, Aston and Salford, all of which improved by considerably higher than the average 0.4 of a grade estimated by the Higher Education Funding Council for England.

Even more striking were those institutions that boosted averages while maintaining the percentage of research active staff.

Goldsmiths College, bottom of the old university sector in 1989 and well down in 1992, has a fair claim to the most spectacular performance in the exercise. It has risen by almost an entire grade in four years while submitting almost exactly the same proportion of staff. Greenwich, Hertfordshire and Thames Valley were among new universities which produced strikingly higher averages while maintaining submission percentages.

Among the research elite the London School of Economics and Warwick joined Cambridge in submitting 98 per cent of staff. Birkbeck's decline from 13th to 32rd place among universities may in part be due to a rise from 87 to 93 per cent of staff assessed.

But the unhappiest institutions are likely to be those that cut the numbers submitting and still fell in the rankings. Exeter suffered an absolute decline in its grade average in spite of submitting 81 per cent of staff against 88 per cent in 1992. Coventry, among the leading new universities in 1992, fell sharply. City and Aberystwyth are now, Lampeter apart, the old universities with the lowest weighted average, in spite of cutting their submission rates.

The league tables show the old order still reigning - Oxford and Cambridge leading parts of London with a mix of 1960s creations at the head of the pursuers. Bath was the conspicuous riser in this group, up from 20th to 7th while putting in 90 per cent of its staff, close to the norm for this group.

There is still a clear division between old and new universities, although Sheffield Hallam, top ex-polytechnic, as in 1992, has squeezed just ahead of Lampeter.

But there was a reshuffle among the other leaders in the class of 1992. Kingston and Coventry fell back while Westminster, Nottingham Trent, Greenwich and Hertfordshire made substantial advances to complete the top five in this sector.

Subject relativities have to be treated with care. The peer-group principle means that outcomes reflect a discipline's estimate of its own research standing rather than an objective comparison with other subjects. Biochemistry, only 17th in the 1992 ranking, emerges as having either the strongest research or the greatest self-regard on this basis.

Twenty of the 69 subject areas rated themselves at an average of 4 - meaning national excellence in all sub-areas - or higher. Other areas showing much higher scores than last time include mineral and mining engineering, anatomy, French and metallurgy and materials. Anthropology, top in 1992, ranks only 12th.